# TC_001 – Modularity gate regression

**ID:** TC_001
**Title:** Modularity gate regression
**Type:** Regression / regime validation

---

## 1. Purpose

Ensure that changes to the instance generator, bounds or configuration do **not** break the
modularity-based gate used to classify or filter instances by community structure.

This test case is linked to the regression test in `tests/test_modularity_gate_regression.py`.

---

## 2. Preconditions

- The project is installed (via Poetry or a plain virtualenv).
- Synthetic instances have been generated according to `configs/instances_to_generate.yaml`.
- Python tests can be executed (see `docs/testing/TESTING.md`).

---

## 3. Procedure

1. Ensure the development environment is set up (Poetry recommended):

   ```bash
   cd ~/MPP
   poetry install --with dev -E metrics
  1. Run the modularity gate regression test:

bash poetry run pytest tests/test_modularity_gate_regression.py

  1. Observe the output, paying attention to:

  2. any failures related to modularity thresholds or regime classification;

  3. any new warnings that indicate borderline cases.

4. Expected results

  • All tests in test_modularity_gate_regression.py pass.
  • Reported modularity values stay within the expected ranges encoded in the test and in the supporting specifications (e.g. specs/bounds.json).

If a failure occurs, the change that triggered it (in the generator, bounds or configuration) must be reviewed. The test expectations should only be updated when there is a deliberate and documented change to the experimental design.

  • Code / tests

  • tests/test_modularity_gate_regression.py

  • Specs / configuration

  • specs/bounds.json (where applicable)

  • configs/instances_to_generate.yaml

  • Documentation

  • docs/reports/01_instance_generator.md

  • docs/testing/TRACEABILITY.md (section on modularity gate)